7482
J. Agric. Food Chem. 2010, 58, 7482–7487 DOI:10.1021/jf101752r
Analysis of Daphnane Orthoesters in Poisonous Australian Pimelea Species by Liquid Chromatography-Tandem Mass Spectrometry SHARON CHOW, MARY T. FLETCHER,* AND ROSS A. MCKENZIE Queensland Primary Industries and Fisheries, Animal Research Institute, Locked Mail Bag No. 4, Moorooka QLD 4105, Australia
Cattle grazing in arid rangelands of Australia suffer periodic extensive and serious poisoning by the plant species Pimelea trichostachya, P. simplex, and P. elongata. Pimelea poisoning (also known as St. George disease and Marree disease) has been attributed to the presence of the diterpenoid orthoester simplexin in these species. However, literature relating to previous studies is complicated by taxonomic revisions, and the presence of simplexin has not previously been verified in all currently recognized taxa capable of inducing pimelea poisoning syndrome, with no previous chemical studies of P. trichostachya (as currently classified) or P. simplex subsp. continua. We report here the isolation of simplexin from P. trichostachya and the development of a liquid chromatography-mass spectrometry/mass spectrometry (LC-MS/MS) method to measure simplexin concentrations in pimelea plant material. Simplexin was quantified by positive-ion atmospheric pressure chemical ionization (APCI) LC-MS/MS with selected reaction monitoring (SRM) of the m/z 533.3 > 253.3 transition. LC-MS/MS analysis of the four poisonous taxa P. trichostachya, P. elongata, P. simplex subsp. continua, and P. simplex subsp. simplex showed similar profiles with simplexin as the major diterpenoid ester component in all four taxa accompanied by varying amounts of related orthoesters. Similar analyses of P. decora, P. haematostachya, and P. microcephala also demonstrated the presence of simplexin in these species but at far lower concentrations, consistent with the limited reports of stock poisoning associated with these species. The less common, shrubby species P. penicillaris contained simplexin at up to 55 mg/kg dry weight and would be expected to cause poisoning if animals consumed sufficient plant material. KEYWORDS: Liquid chromatography-mass spectrometry; simplexin; daphnane orthoester; Pimelea
INTRODUCTION
Cattle grazing in arid rangelands of Australia suffer periodic extensive and serious poisoning by the plant species Pimelea trichostachya, P. simplex and P. elongata (1). These are wintergrowing annual herbs, endemic to Australia and occurring across much of its inland grazing regions (2, 3). Their local abundance depends on variations in climatic and environmental conditions (4). Pimelea poisoning of cattle, known as St. George disease or Marree disease, is a unique syndrome combining pulmonary hypertension causing right ventricular dilation and circulatory failure with subsequent edema, anemia from hypervolemia, chronic diarrhea, and dilation of hepatic sinusoids (peliosis hepatis) (5-8). Horses are rarely affected (9). The impacts of pimelea poisoning vary with seasonal conditions (4), but in severe years, the disease has been estimated to cost the Australian beef industry $AUD50 million AUD in terms of production losses, stock deaths, control measures, and extra management work for cattle producers. *To whom correspondence should be addressed. Tel: 61 7 3362 9426. Fax: 61 7 3362 9429. E-mail:
[email protected].
pubs.acs.org/JAFC
Published on Web 05/27/2010
Feeding trials with cattle in the 1960s and 1970s established Pimelea species as the cause of this syndrome (8, 10-16). Subsequent taxonomic revisions of Pimelea species (2, 17, 18) have caused us to re-examine the lodged herbarium voucher specimens to determine the current botanical identity of plant material employed in these experimental studies. Before revision, the taxon P. trichostachya broadly included P. elongata, P. simplex subsp. continua, and the current P. trichostachya in its restricted sense. For example, five voucher specimens lodged by Clark and cited in the literature as P. trichostachya (8), have now been determined by Queensland Herbarium to be individual collections of all three of these taxa (Table 1). On the basis of the redetermination of vouchers, currently recognized taxa that have previously experimentallyinducedpimelea poisoningareP.trichostachya(8,13,19), P. elongata (8, 13, 15, 16), P. simplex subsp. continua (8), and P. simplex subsp. simplex (14). Full details of taxa reassignments relating to these earlier investigations are shown in Table 1 and demonstrate the value in lodging herbarium voucher specimens and citing corresponding voucher numbers in scientific publications. Voucher numbers were not cited for several investigations conducted in New South Wales, and the relationship with existing
© 2010 American Chemical Society
Article
J. Agric. Food Chem., Vol. 58, No. 12, 2010
7483
Table 1. Current Determinations of Voucher Specimens Lodged for Previous Pimelea Toxicity and Toxin Investigations reference
collection site
herbarium vouchera
species as cited in original reference text
current determination of voucher specimensb
Feeding Trial Research 16 c
Bourke NSW
P. simplex
(NSW412853)
P. elongata
15 c
Bourke NSW
P. simplex
(NSW590111 or NSW412853)
P. elongata
14
‘Kulkyne’ NSW
P. simplex
(NSW651609)
P. simplex subsp. simplex
8
3 localities in western Qld
P. trichostachya P. trichostachya P. trichostachya P. trichostachya P. trichostachya
BRI099670 = AQ86121 BRI101503 = AQ97842 BRI101606 = AQ170241 BRI101502 = AQ86134 BRI109643 = AQ170221
P. trichostachya P. trichostachya P. simplex subsp. continua P. elongata P. elongata
13
‘Macwood’ Qld
P. trichostachya P. trichostachya P. trichostachya P. continua
BRI129152 = AQ1682 BRI129169 = AQ1670d BRI129170 = AQ1673 BRI129168 = AQ1653
P. trichostachya P. trichostachya P. trichostachya P. elongata
19
St. George Qld
P. trichostachya
BRI129169 = AQ1670
P. trichostachya
Simplexin Isolation Reports e
e
20 , 24 , 22
‘Kulkyne’ NSW ‘Willoring’ NSW ‘Belvedere’ NSW
P. simplex P. simplex P. simplex
(NSW651609) (NSW651733) (NSW590111)
P. simplex subsp. simplex P. simplex subsp. simplex P. elongata
23 f
‘Belvedere’ NSW ‘Kulkyne’ NSW ‘Belvedere’ NSW
P. simplex P. simplex P. trichostachya Form B
(NSW590111) (NSW651609) (NSW251889)
P. elongata P. simplex subsp. simplex P. elongata
a Voucher numbers are as cited in original text, and numbers in parentheses were submitted from collection property/author or co-workers around the appropriate time period (information from NSW Herbarium), but actual voucher numbers are not cited in publications. b Current determination from re-examination of herbarium voucher specimens. c J. E. Cantello lodged NSW412853 and NSW590111 from ‘Belvedere’ in 1967 and 1971 for feeding trials/experiments at Glenfield Veterinary Research Station, and H. B. Roberts continued this work. d Collection listed by Kelly (13) as “BRI129160” is believed to be a typographical error as this number is not consistent with the Queensland Herbarium collection, which shows “BRI129169” as a feeding trial sample from W. R. Kelly. e Isolation of simplexin from P. simplex collected by H. B. Roberts in the Bourke area is presumed to relate to herbarium samples from ‘Kulkyne’, ‘Willoring’ and/or ‘Belvedere’ as cited by Hafez et al. (22). f Authors acknowledged plant collections by H. B. Roberts and T. J. McClure.
herbarium specimens (Table 1) could only be construed from interrogation of details recorded by New South Wales Herbarium (such as name of collector, original determination, collection date, and reason for collection). The toxin responsible for this poisoning syndrome has been identified as the novel daphnane orthoester simplexin (20). It was reported that intravenous inoculation of cattle with isolated pure simplexin (4 μg/kg body weight) produced a 3-fold increase in pulmonary arterial pressure within 100 s, while oral dosing reproduced characteristic signs of pimelea poisoning, including inappetence, weight loss, jugular distention, and submandibular edema (20). Daphnane orthoesters are an unusual class of compounds found only in the plant families of Thymelaeaceae and Euphorbiaceae, and the occurrence and biological activity of such compounds have recently been reviewed (21). Simplexin has been isolated previously from P. simplex subsp. simplex (cited originally as P. simplex) (20, 22-24) and also from P. elongata (cited originally as P. trichostachya Form B and P. simplex) (23) (Table 1). A number of compounds of related structure have also been identified from these species (22, 23, 25). P. trichostachya as currently classified and P. simplex subsp. continua have not been chemically examined previously, although these species have been assumed to contain simplexin due to their major contributions to the occurrence of pimelea poisoning of cattle in the field. Additionally, P. trichostachya has recently been reported to cause this syndrome in horses (9). This study was undertaken to develop a liquid chromatography-mass spectrometry/mass spectrometry (LC-MS/MS) method to determine whether simplexin was indeed present in
P. trichostachya and P. simplex subsp. continua, and to determine concentrations of simplexin in each of the four poisonous Pimelea taxa that commonly induce the syndrome in the field. Several other Pimelea species also occur in inland grazing areas of Australia, and some have on occasion been suspected of inducing the pimelea poisoning syndrome (1). In this study, simplexin concentrations have also been determined in samples of suspect species, P. decora, P. haematostachya, and P. microcephala, and also in P. penicillaris. These four species have not previously been chemically examined. MATERIALS AND METHODS Plant Material. Collections of P. trichostachya, P. elongata, P. simplex subsp. continua, and P. simplex subsp. simplex were made of new growth and flowering tops (those parts of plants considered most likely to be eaten by grazing livestock), and were a composite of 10 or more plants collected at each site at various locations in western Queensland and New South Wales. Individual samples of P. decora, P. haematostachya, P. microcephala, and P. penicillaris were also collected for comparison. The nature of collection sites and global positioning system (GPS) coordinates were recorded. Samples (200 g-1 kg) were placed in paper bags. A separate sample was pressed between absorbent paper for botanical identification. Plant identifications were confirmed by the Queensland Herbarium, and botanical specimens from all batches collected were incorporated into their permanent collection as vouchers against any future taxonomic changes. Field-collected plant material was transported to the laboratory, airdrying completed, milled, and stored frozen prior to analysis. Isolation of Simplexin from P. trichostachya. Simplexin standard was obtained by extraction of a milled bulk stem and root sample of P. trichostachya (AQ751555, collected November 2006) from a property near Mitchell, Queensland. The plant sample (40 g) was soaked in 90%
7484
J. Agric. Food Chem., Vol. 58, No. 12, 2010
Chow et al.
Table 2. Molecular Ions and Selected MS/MS Fragment Ions of Daphnane Orthoesters 1-7 compound
[M þ H]þ
major fragment ions (relative abundance, %)
dehydrosimplexin (1) acetoxysimplexin (2) dodecenyl analogue (3) simplexin (4) dehydrohuratoxin (5) acetoxyhuratoxin (6) huratoxin (7)
531 591 559 533 583 643 585
531 (85), 361 (7), 343 (14), 325 (27), 307 (42), 297 (25), 279 (47), 267 (61), 253 (100), 153 (53, R1COþ)a 591 (16), 401 (9), 383 (11), 365 (17), 359 (19), 341 (55), 323 (100), 295 (69), 277 (29), 269 (60), 155 (12, R1COþ)a 559 (100), 361 (8), 343 (20), 325 (30), 307 (48), 297 (25), 279 (27), 267 (74), 253 (83), 181 (29, R1COþ)a 533 (32), 361 (5), 343 (14), 325 (29), 307 (29), 297 (26), 279 (26), 267 (57), 253 (100), 155 (10, R1COþ)a 583 (100), 361 (3), 343 (3), 325 (8), 307 (7), 279 (7), 267 (23), 253 (32), 205 (56, R1COþ)a 643 (59), 401 (10), 365 (6), 359 (7), 341 (18), 323 (19), 295 (21), 277 (21), 269 (12), 207 (100, R1COþ)a 585 (100), 361 (3), 343 (7), 325 (15), 307 (12), 297 (9), 279 (15), 267 (32), 253 (41), 207 (40, R1COþ)a
a
R1 represents the fatty acid chain of the orthoester in each compound (Figure 1).
methanol overnight (200 mL). The mixture was filtered, and the plant material was rinsed with fresh methanol (50 mL). The combined organic extracts were concentrated under reduced pressure and the residue partitioned between 50% saturated brine solution (50 mL) and dichloromethane (3 50 mL). After drying over anhydrous sodium sulfate and solvent removal, the green residue was partitioned between hexane (50 mL) and acetonitrile (3 20 mL). The acetonitrile layers were combined and concentrated under reduced pressure. The residue was dissolved in minimum amount of dichloromethane and loaded onto four Strata SI-1 Silica SPE cartridges (Phenomenex, Sydney, Australia, 2 g), eluting with dichloromethane and then 1% methanol in dichloromethane. The fractions containing the desired toxin (guided by TLC visualization with ammonium metavanadate reagent (23) and confirmed by LC-MS/MS) were concentrated and purified by reverse phase high performance liquid chromatography (HPLC) on a 250 15 mm i.d., 5 μm, Luna C18(2) column, with a 10 mm 10 mm i.d. guard column of the same material (Phenomenex, Sydney, Australia) at 25 °C and eluted isocratically with 95% methanol. The HPLC was equipped with a diode array detector and low temperature evaporative light scattering detector (ELSD). The identity of the isolated simplexin (6 mg from 180 g of dried P. trichostachya stems and roots) was confirmed by 1H and 13C NMR data comparisons with literature values (23) and shown to be g95% in purity by HPLC-ELSD. Additional simplexin was also isolated by a similar procedure from P. elongata (AQ751686, collected July 2007) from Bollon, Queensland. Analytical Sample Preparation. Milled dried plant samples (0.5 g) were extracted by shaking overnight in methanol (80%, 20 mL). A portion of the centrifuged extract (2 mL) was evaporated under a stream of nitrogen with the residue taken up in dichloromethane (4 mL) and distilled water (1 mL), and washed with brine (5 mL). The organic layer was removed, and the aqueous layer was re-extracted with dichloromethane (2 4 mL). After drying the combined organic layers with anhydrous sodium sulfate, the solvent was evaporated. The residue was partitioned between acetonitrile (4 mL) and hexane (10 mL). The hexane layer was washed with fresh acetonitrile (2 mL). Solvent was evaporated from the combined acetonitrile extracts, and the residue was taken up in methanol (8 mL) for LC-MS/MS analysis. LC-MS/MS Equipment and Conditions. Samples were analyzed using a Waters 2777C Sample Manager and the liquid chromatographic separations were carried out on a Waters 1525μ Binary HPLC pump system. LC separations were performed using a 2.1 100 mm i.d., 3.5 μm, Sunfire C18 column (Waters, Sydney, Australia) at 30 °C with a flow rate of 0.2 mL/min. The mobile phase was a mixture of (A) methanol with 0.1% formic acid (v/v) and (B) water with 0.1% formic acid (v/v), with a gradient as follows: t0, 90% A; t15, 100% A; t25, 100% A; t26, 90% A;. t30, 90% A. MS/MS detection was made using a Quattro Premier Micromass triple quadrupole mass spectrometer, equipped with an atmospheric pressure chemical ionization (APCI) source used in positive mode. The mass spectrometer response was tuned with a simplexin solution. The capillary voltage was 3.0 kV; the desolvation and cone gases of nitrogen flow were set at 450 L/h and 47 L/h, respectively. The desolvation and source temperatures was set at 450 and 120 °C, respectively. Argon was used as collision gas for MS/MS with a flow rate of 0.3 mL/min, and the collision energy was set at 16 eV and cone voltage at 20 V. Isolated simplexin (g95% pure) was used as an external standard in LC-MS/ MS analysis with standard solutions prepared in methanol (0.5-5.5 mg/L). RESULTS AND DISCUSSION
Isolation of Simplexin from P. trichostachya. The crude plant extract of P. trichostachya obtained from the liquid-liquid partition
Figure 1. Structures of simplexin (4) and related compounds detected in the Pimelea plant species.
consisted of a complex mixture of components in which simplexin and related diterpenoid esters were present at only low levels. TLC visualization with ammonium metavanadate reagent (23) was used to guide SPE fractionation of the extract, and LC-MS/ MS analysis was used to confirm the presence of simplexin in the collected fractions. On the basis of the characteristic UV band of simplexin (242 nm), the toxin was separated by repetitive semipreparative HPLC, and the purity of the isolated material was assayed at g95% from ELSD and NMR analyses. The structure of the isolated simplexin was confirmed by mass spectrometry and also a comparison of 1H and 13C NMR data with literature values (23). This is the first reported isolation of simplexin from P. trichostachya (as currently classified) and further confirms the role of simplexin in poisonings of stock by this plant species. LC-MS/MS Fragmentation of Simplexin and Analogues. LC-MS/MS fragmentation of the protonated molecular ion of simplexin [M þ H]þ m/z 533 in positive-ion APCI mode showed loss of its fatty acid (C9H19COOH) to a m/z 361 fragment with further sequential losses of 18 Da (H2O) to m/z 343, 325, and 307 (Table 2). A small C9H19COþ ion at m/z 155 was observed. Using the selected reaction monitoring (SRM) program on the LC-MS/ MS, the most dominant daughter ion observed at m/z 253 was chosen as the transition (m/z 533.3 > 253.3) for quantification, and the next dominant daughter ion was selected as the transition (m/z 533.3>267.2) for verification (Table 3). The efficiency of the ionization and the detection of the precursor and fragments ions were optimized by adjusting collision energy and cone voltage (16 eV and 20 V, respectively). The initial LC-MS/MS investigations of the crude plant extract of P. trichostachya, P. elongata, P. simplex subsp. continua, and P. simplex subsp. simplex revealed the presence of other minor components as well as simplexin. The mass fragmentation pattern of several of these components reflected a certain degree of
Article
J. Agric. Food Chem., Vol. 58, No. 12, 2010
7485
Figure 2. LC-MS/MS analysis of Pimelea trichostachya (AQ751829). (A) Individual SRM chromatograms for each of the daphnane orthoesters 1-7. (B) Overlaid SRM chromatograms.
structural similarity to simplexin, and mass spectral data for all seven compounds (including simplexin) are listed in retention order, 1-7, in Table 2. From the mass spectrometric fragmentation, it was presumed that these compounds possessed the same carbon skeleton but with different fatty acid and side chains attached (Figure 1). Compounds 1, 3, 5, and 7 have MS/MS spectra very similar to those of simplexin (4), including the presence of dominant daughter ions at m/z 267 and m/z 253 as shown in Table 2. The R1COþ ions, from the fragmentation of the fatty acids, enabled the assignments of the alkyl fatty acids to structures 1, 3, 5, and 7 (Figure 1). Two other minor components 2 and 6 had protonated molecular ions [M þ H]þ of m/z 591 and m/z 643, with mass fragmentation somewhat similar to that seen in simplexin (4) and huratoxin (7), correspond to acetoxy-derivatives of simplexin and huratoxin (increments of 58 Da., Table 2). These compounds were deduced to be acetoxysimplexin (2) and acetoxyhuratoxin (6). Huratoxin (7) and acetoxyhuratoxin (6) have previously been isolated from P. simplex (23) and acetoxysimplexin (also known as gnidiglaucin) (2) from Gnidia glaucus (26). Development and Validation of the LC-MS/MS Method for Simplexin Analysis. Simplexin (4) is unsuitable for GC-MS analysis and does not have a strong UV chromophore for HPLC-UV analysis; thus, the described LC-MS/MS method provides the first sensitive analysis method for this toxin in Pimelea taxa. Such a method was required as part of our study to provide a better understanding of the incidence of pimelea poisoning by relating observed field toxicity to measured simplexin concentrations in the various Pimelea species. Simplexin (4) is a moderately polar molecule, and a solvent partitioning procedure was developed to provide crude pimelea plant extracts containing simplexin (and related minor components) free from both water-soluble and hydrocarbon components. Reverse phase HPLC on a C18 column eluting with a methanol/ water (containing 0.1% formic acid v/v) gradient then provided the separation of components 1-7 (Figure 2). Concentrations of simplexin (4) in plant extracts were determined from the intensity of the positive-ion APCI SRM transition m/z 533.3> 253.3 and verified by a secondary transition (m/z 533.3> 267.2). The presence of other minor components 1-3 and 5-7 in plant extracts were determined by measurement of similar SRM transitions (and confirmatory SRM transitions), as listed in Table 3. Absolute levels of these compounds were not determined as adequate standards have not been separated for these minor components. Under the described HPLC conditions, all
Table 3. Selected Reaction Monitoring (SRM) Transitions for Daphnane Orthoesters 1-7 compound
RT (min)
quantification SRM
confirmatory SRM
dehydrosimplexin (1) acetoxysimplexin (2) dodecenyl analogue (3) simplexin (4) dehydrohuratoxin (5) acetoxyhuratoxin (6) huratoxin (7)
4.4 4.7 5.7 5.8 6.3 6.4 7.3
531.5 > 253.3 591.5 > 323.1 559.5 > 253.3 533.3 > 253.3 583.5 > 253.3 643.3 > 207.2 585.5 > 253.3
531.5 > 267.2 591.5 > 295.3 559.5 > 267.2 533.3 > 267.2 583.5 > 267.2 643.3 > 277.2 585.5 > 267.2
seven components eluted with retention times between 4.4 and 7.3 min (Table 3). Simplexin (4) in plant extracts from individual field samples was quantitated against a four point calibration curve of simplexin (standards were prepared in methanol) over the concentration range of 0.5-5.5 μg/mL. Plant extracts with simplexin concentrations outside this range were further diluted to bring them back into range. Typically, calibration curves had R2 values in excess of 0.96. Reproducibility of analysis was demonstrated by replicate analysis. For example, five replicates of a single sample of flowering P. elongata (AQ751686) had a mean simplexin concentration of 275.7 mg/kg DW with a standard deviation of 7.6 (RSD 2.76%). To estimate the recovery efficiency of the extraction method, authenticated (NMR, MS) simplexin (4) was spiked into dried milled plant material from P. simplex subsp. continua (AQ751854). This plant material contained minimal natural simplexin (average analysis of 24.8 mg simplexin/kg dry weight), was spiked at concentrations equivalent to approximately 112, 224, 384, and 880 mg simplexin/kg dry weight, and extracted by the described procedure to afford extracts with theoretical concentrations of 0.7, 1.4, 2.4, and 5.5 μg/mL. When compared with simplexin standard solutions in methanol and allowing for natural simplexin level, these spiking experiments had recoveries of 168, 140, 121, and 92% respectively, and demonstrated that the LC-MS/MS response for simplexin was somewhat enhanced by the presence of matrix components particularly at low simplexin levels. When compared with simplexin standard solution in the plant extract matrix, calculated recoveries were 95, 87, 104, and 92% respectively. Composition of Pimelea Taxa. LC-MS/MS analysis of the four poisonous taxa P. trichostachya, P. elongata, P. simplex subsp. continua, and P. simplex subsp. simplex showed similar composition with simplexin (4) as the major diterpenoid ester component
7486
J. Agric. Food Chem., Vol. 58, No. 12, 2010
Chow et al.
Table 4. Chemical Composition of Green Flowering Samples of Pimelea Taxa
Pimelea taxon
location (nearest town)
simplexin content other (mg/kg diterpenoids DW) identified
P. decora Domin
Julia Creek
